Kritika Maheshwari asks:

Is it morally wrong to expose others to risk of harm even if no one ever ends up harmed?

This is one of the central questions I tackle in my (2019) book Falibilidad y Normatividad (Falibility and Norms), since I know many people cannot read Spanish, here is a very brief recount of my answer in English: 

The questions seems difficult because it is equivocal, for there is not one  such thing as being right. We use normative language like ‘right’ and ‘wrong’ for many purposes – to talk about what we should do from a first person perspective and to talk about how to judge other people’s actions from a third person perspectives (singular and plural), about who to praise and who to blame, about how to deal with the consequences of our actions and about how to assign responsibilities, etc. – and once we disentangle them, the answer to the question becomes no longer puzzling and may even be trivial. For example, if what we mean by ‘morally wrong’ just means whether someone should be held responsible for the harm incurred, then the trivial answer is “no”: if no one ends up harmed, then none needs to be held responsible, etc,

Furthermore, the reason why the question seems puzzling is because it is formulated as a dilemma that uses, on one horn, language that favours one sort of interpretation and on the other language that favours another, incompatible interpretation. Talk of ‘exposing others to risks’ is relevant when the question is about how to act in situations of uncertainty; talk of whether someone ‘ends up harmed’ or not, in contrast, is relevant only when the decision has been made, the act performed and we are in a position to know whether harm was incurred or not and thus is irrelevant to the original question about how to act in situations of uncertainty and instead is relevant to other, different questions, related to how to deal with the consequences of our action or other’s. For these later questions, on the other hand, talk of risk is irrelevant, for the consequences have already unfolded. Thus, when we try to interpret ‘morally wrong’ in Maheshwari’s question we are left with the mistaken impulse to think that there is a more general way of understanding right and wrong for which both horns of the dilemma are relevant, but ultimately there is not.